There are a host of applications that call for ultrahigh resolution frequency measurement and synthesis ranging from optical metrology, optical frequency standards, ultrahigh resolution atomic spectroscopy, and ultrahigh resolution optical frequency and time domain multiplexing, among others. The utility of optical frequency measurement and synthesis in such applications has been limited on the one hand by the inability of the heretofore known techniques to provide frequency measurements of optical frequency sources with the same resolution as the fractional stability of the optical frequency sources, and has been limited on the other hand by the inability to provide ultrastable high resolution optical frequencies selectively at wavelengths of practical interest.
High-resolution, high-accuracy spectroscopy of laser-cooled and trapped single ions is expected to yield a resolution on the order of one part in 10.sup.18, as reported in an article entitled "Laser-Cooling Limits and Single Ion Spectroscopy", by Wineland et al., Physical Review A36, 2220 (1987), incorporated herein by reference.
Two-photon spectroscopy of an atomic fountain of neutral hydrogen atoms is expected to yield a resolution on the order of one part in 10.sup.15, as reported in an article entitled "Ultrahigh-Resolution Two-Photon Optical Ramsey Spectroscopy of an Atomic Fountain", by Beausoleil and Hansch, Physical Review A33, 1661 (1986), incorporated herein by reference.
The highest resolution optical metrology techniques heretofore, namely those based on optical interferometric principles, however, have only been able to measure such optical frequencies with a resolution and accuracy several orders of magnitude less than the resolution with which the lines are to be provided. The limitation on measuring these lines with a resolution several orders of magnitude less than their linewidths imposed by the heretofore known optical metrology techniques has limited the exploitation of these and other sources of ultrahigh stability optical frequencies. In order to meet future requirements of 10.sup.-15, or better, resolution, non-interferometric direct frequency measurements are necessary.
A frequency synthesis chain has been demonstrated to compare the 633 nm He-Ne laser stabilized on a molecular iodine line to the primary frequency standard, the 9.2 GHz cesium clock, as reported in an article entitled "Direct Frequency Measurement of the I.sub.2 -Stabilized He-Ne 473-THz (633-nm) Laser", by Jennings et al, appearing at Optics Letters 8, 136 (1983), incorporated herein by reference. The utility of the frequency link, however, depends on and is thus limited by the coincidence of certain harmonics of laser and klystron sources.
Non-resonant interaction in nonlinear crystals, as suggested in an article entitled "Novel Optical Frequency Divider and Synthesizer", by McIntyre and Hansch, appearing at Technical Digest, 1988 Annual Meeting of the Optical Society of America, p. 131, incorporated herein by reference, uses sum and difference frequency mixing in nonlinear crystals for frequency division and synthesis. This approach, however, has low efficiency and therefore has an undesirably low S/N ratio.